Primary mediastinal large B-cell lymphoma

What every physician needs to know:

Primary mediastinal (thymic) large B-cell lymphoma (PMBCL) was recognized as a distinct subtype of diffuse large B-cell lymphoma (DLBCL) in the Revised European American Lymphoma classification in 1994, and more recently included in the World Health Organisation (WHO) classification, based on unique clinical, pathologic and genetic features.

By definition, patients have an anterior mediastinal mass, and the disease usually occurs in young adults with a median age of approximately 35 years and a female predominance (twice as many women as men).

Are you sure your patient has primary mediastinal large B-cell lymphoma? What should you expect to find?

Patients present with an anterior, often bulky mediastinal mass, and may have associated respiratory symptoms. Superior vena cava (SVC) syndrome can occur with facial swelling, dyspnea, headache, neck vein distention, and occasionally, thrombosis. Patients typically have bulky, stage I or II disease and B symptoms (fever, night sweats, weight loss) at diagnosis. Intrathoracic extension into the lung, chest wall, pericardial and pleural spaces is common, however, extrathoracic disease, including bone marrow involvement, at presentation is rare. Further spread to peripheral lymph nodes is infrequent, although localized extension to the supraclavicular nodes can occur. Unusual extranodal sites can be involved at relapse such as the liver, kidneys, and central nervous system.

Continue Reading

PMBCL tumors have a diffuse growth pattern of large cells with clear cytoplasm and it is commonly associated with fibrosis, which tends to compartmentalize the tumor.

The following immunophenotypic markers are characteristically positive: CD45 (leukocyte common antigen) and the B-cell markers CD19, CD20, CD22 and PAX5. However, unlike other B-cell lymphomas, they can lack surface or cytoplasmic immunoglobulin (Ig), despite expression of the Ig co-receptor, CD79a. CD30 is often present, but is usually weak and heterogeneous, in comparison to classical Hodgkin lymphoma (CHL). However, CD15 is absent.

Most cases are CD10 negative, but have multiple myeloma-1/interferon regulatory factor-4 (MUM1/IRF4) positive phenotype with variable B-cell lymphoma 6 (BCL6) expression. Tumour cells are CD21-negative, but can be CD23-positive in up to 80% of cases, consistent with possible thymic B-cell derivation. Expression of MAL antigen, a lipid raft component, is another unique characteristic feature of PMBCL, not found in other DLBCL. Lack of HLA class I and/or class II molecules has been reported.

Thus, the diagnosis should be based on the typical clinical presentation, in addition to the distinctive clinico-pathologic features.

Beware of other conditions that can mimic primary mediastinal large B-cell lymphoma:

PMBCL can mimic a carcinoma or thymoma due to the fibrosis. Multilobated nuclei which may resemble Hodgkin Reed-Sternberg cells (HRS) can be seen. The diagnosis can be challenging when biopsies are small and in the presence of sclerosis, necrosis, and/or cellular crush artifacts. Remnants of thymic epithelial structures may be observed, particularly on large biopsies or surgical specimen, which are often better highlighted using cytokeratin staining. A fine needle aspirate is never adequate for diagnosis as both morphology and immunophenotyping are essential for diagnostic accuracy.

PMBCL can also be difficult to distinguish from nodular sclerosis classical Hodgkin lymphoma (NSCHL). Both entities occur in young adults, and anterior mediastinal involvement is common in NSCHL. Recent gene expression profiling data suggests there is also a molecular overlap between NSCHL and PMBCL. Further, there is a new entity in the WHO lymphoma classification of mediastinal gray zone lymphomas (MGZL) that combines morphologic and phenotypic features of both NSCHL and PMBCL, and can be a diagnostic challenge to pathologists.

These ‘gray zone lymphomas’ are defined in the WHO as B-cell lymphoma, unclassifiable, with features intermediate between diffuse large B-cell lymphoma and classical Hodgkin lymphoma and often present in younger males (age 20 to 40 years) with anterior mediastinal involvement. The distinction is important as MZGL appears to have a worse prognosis.

MGZL typically are composed of confluent sheet-like growths of pleomorphic tumour cells in diffusely fibrotic stroma. The cells are usually larger and more pleomorphic than in PMBCL. Different areas of the tumor can show variations in cytological appearance, with some areas more closely resembling CHL and others appearing more like DLBCL. There is usually a sparse inflammatory infiltrate. The immunophenotype has characteristics of CHL and PMBCL. The cells usually express CD45 and in contrast to CHL, the B-cell program is preserved, but Hodgkin type markers such as CD30, CD15 are aberrantly expressed. B-cell markers CD20 and CD79a are often positive, but surface Ig is absent. MAL is often seen, supporting a link to PMBCL.

Another consideration in the differential diagnosis is “classical” DLBCL, associated with secondary mediastinal invasion which most likely originates from mediastinal or hilar lymph nodes. In contrast to PMBCL, extrathoracic disease is much more common and tumor sclerosis is absent. Further, mediastinal involvement is not the prominent site of disease. Gene expression profiling studies have also shown that PMBCL has a distinct profile compared to DLBCL, with a number of uniquely expressed genes. PMBCLs are also characterized by the presence of both TNF- receptor-associated factor 1 (TRAF1) and nuclear c-REL which have has been shown to be highly specific for the diagnosis of PMBCL, compared to other DLBCLs.

Which individuals are most at risk for developing primary mediastinal large B-cell lymphoma:

There is no known cause of PMBCL, but it is seen most commonly in young adult women.

Epstein-Barr virus, which can be present in CHL tumors, has not been found in PMBCL.

What laboratory studies should you order to help make the diagnosis and how should you interpret the results?

Pathologic review by an expert hematopathologist is essential for the diagnosis of PMBCL, and to ensure the biopsy was adequate. Although laboratory parameters such as LDH (lactate dehydrogenase) are used for prognostication, there is no laboratory test that can diagnose any lymphoma.

A bone marrow aspirate and biopsy should also be performed to complete staging, although it is rarely positive and the diagnosis should be re-considered in this setting.

What imaging studies (if any) will be helpful in making or excluding the diagnosis of primary mediastinal large B-cell lymphoma?

Computed tomography (CT) imaging from neck to pelvis are performed in PMBCL for staging. By definition, the mediastinum should be the predominant site of disease; extrathoracic disease, if present, should be minimal. The revised response criteria for lymphoma supports that fluorodeoxyglucose positron emission tomography (FDG-PET) should be used at the end of therapy to assess response. For full interpretation of a post-treatment FDG-PET scan, a staging FDG-PET should also be ideally performed. In contrast to CT imaging, FDG-PET is a functional imaging tool that can distinguish between viable tumor and necrosis or fibrosis in a residual mass.

If you decide the patient has primary mediastinal large B-cell lymphoma, what therapies should you initiate immediately?

PMBCL is an aggressive lymphoma and once the diagnosis and staging studies have been completed, therapy should be started as soon as possible. If there are symptoms of SVC syndrome (see above), steroids can be considered as chemotherapy is being organized.

Radiation should be avoided, with the possible exception of one or two doses to alleviate severe symptoms, prior to chemotherapy, as it may delay definitive curative therapy and increase toxicity.

More definitive therapies?

The standard therapy in PMBCL is considered to be R-CHOP (rituximab/cyclophosphamide/doxorubicin hydrochloride/vincristine/prednisolone), although there has been some debate over what the best back-bone chemotherapy is in PMBCL, particularly in the pre-rituximab treatment era. Some retrospective studies have suggested that more dose intensive regimens such as MACOPB (methotrexate/cytarabine/Cyclophosphamide/Oncovin/prednisone/bleomycin) or VACOP-B (etoposide/LV/doxorubicin/cyclophosphamide/vincristine/prednisone/bleomycin) may be superior to CHOP (cyclophosphamide, doxorubicin, vincristine, prednisone), but this has not been tested in a randomized trial.

More recently, the addition of rituximab to CHOP chemotherapy has been shown in multiple studies to improve cure rates in DLBCL over CHOP alone. A randomized trial comparing CHOP to R-CHOP has not been performed in PMBCL, due to disease rarity. Thus, the true value of adding rituximab to CHOP in PMBCL is unknown, however, it is likely that the same magnitude of benefit seen in DLBCL will be observed, and this may negate any benefit of dose-intensity.

The effect of intensified regimens has already been reported in comparisons of CHOP to CHOEP (cyclophosphamide, doxorubicin, vincristine, etoposide, prednisone) in young, good prognosis, aggressive lymphoma patients, the majority of which were DLBCL (non-Hodgkin lymphoma-B1 [NHL-B1]). In this pre-rituximab treatment era, patients treated with CHOEP had an improved event free survival (EFS) over CHOP alone. However, with the addition of rituximab, this benefit was no longer apparent, suggesting that rituximab may be the “equalizer” of treatment regimens.

The MINT trial (Mabthera International Trial Group study) compared six cycles of CHOP-like regimens (most commonly CHOP-21 [48%] CHOEP [44%]), with and without rituximab in young, good risk patients defined as age 18 to 60 years, with age adjusted International Prognostic Index 0 or 1 (risk factors: stage III/IV disease, elevated LDH and performance status of greater than two), and reported a superior EFS and overall survival (OS) in the patients who received rituximab. The outcome of the subset of patients with PMBCL patients randomized in this study (n=87) similarly noted an improved 3 year EFS in patients who received rituximab (78%, n=44), compared to those that did not (52%, n=43), (p=0.012) , but OS was not statistically different (89% versus 78%, p=0.158) , which may reflect the small number of patients in each treatment group. The patient numbers were too small to evaluate the added benefit of etoposide, comparing R-CHOEP to R-CHOP. Importantly, in multivariate analysis, treatment with rituximab and the absence of bulky disease were predictors for a favorable outcome. Treatment with rituximab also resulted in a significant reduction of the frequency of progressive disease (2.5% versus 24%, p=0.006). A retrospective study reported the outcome of 76 patients treated with R-CHOP and describe a similar 3 year freedom from progression (FFP) of 81% and a 3 year OS of 89%. Overall, the reported studies would suggest that approximately 20% of patients fail first-line therapy with R-CHOP.

The role of radiotherapy in PMBCL remains ill-defined. A major challenge in the management of PMBCL, is the evaluation of a residual mass post chemotherapy, and whether it represents active lymphoma. There is poor correlation between the size of a residual mass on computerized tomography and risk of relapse, since in many cases, the residual density may represent fibrotic tissue, as opposed to active lymphoma. Consolidative mediastinal radiotherapy is often delivered; however, it is unclear whether this impacts relapse or cure rates, and there is an inherent concern of long-term toxicities of mediastinal radiotherapy, including an increased risk of cardiovascular disease and secondary malignancies, particularly given the young population at risk.

Although some studies have supported that radiotherapy improves EFS, other analyses have demonstrated that chemotherapy alone is effective in many cases, suggesting that radiotherapy is not mandatory in all patients. An analysis evaluating the impact on progression-free survival with a policy recommending routine radiotherapy following primary chemotherapy, failed to demonstrate a benefit however, different chemotherapy regimens were used. The retrospective nature of such analyses, including definitions of response rates is problematic, and randomized studies addressing this question are lacking. Improved identification of patients who may benefit from the addition of radiotherapy is needed.

In the MINT retrospective analysis, approximately half of the patients received radiotherapy according to center guidelines for bulky disease (variable definitions) or extranodal involvement. The authors noted that 30% achieved an improvement of response with radiotherapy, but it is not clear whether this truly converted partial remissions to complete remissions, as PET scanning was not done, and the impact was not correlated with outcome.

Recently, a dose intensive approach using dose adjusted (DA)-EPOCH-R (etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin, rituximab) has shown excellent results in a prospective phase II study of PMBCL, and may obviate the need for radiotherapy. In this protocol, the doxorubicin, vincristine and etoposide are infused over 96 hours based on the premise that continuous drug exposure may enhance cell kill. There is additional pharmacodynamic adjusted dosing of the etoposide, doxorubicin and cyclophosphamide by 20% each, to achieve a neutrophil nadir below 0.5. The results of 26 patients, including some children, treated with DA-EPOCH-R was compared in a non-randomized fashion to 18 PMBCL patients treated with DA-EPOCH and the EFS (p=0.038) and OS (p=0.023) favored DA-EPOCH-R. The results were updated at the 2011 Lugano meeting of 44 patients treated with DA-EPOCH-R and the 3 y EFS and OS of 95% and 100%, respectively. Only two patients received radiotherapy.

These encouraging results await more mature data. Whether these results represent a true superiority of DA-EPOCH-R over R-CHOP or patient selection is unknown. In addition, an on-going US Intergroup study comparing R-CHOP to DA-EPOCH-R without radiotherapy in DLBCL , including PMBCL will be completed in the near future and hopefully will clarify the role of DA-EPOCH-R in these patients. At this time, it is reasonable to consider R-CHOP with or without mediastinal radiotherapy as the standard treatment in PMBCL, akin to the current practice in DLBCL.

The International extranodal study group (IELSG) plan to evaluate the use of PET scanning to guide radiotherapy in PMBCL following a rituximab containing doxorubicin based combination chemotherapy (including R-CHOP and DA-EPOCH-R). If cases are PET negative following chemotherapy, patients will be randomized to either radiotherapy or observation.

What other therapies are helpful for reducing complications?


What should you tell the patient and the family about prognosis?

Overall the cure rate of PMBCL is excellent, and in the MINT and Greek retrospective study, approximately 80% of patients are cured with front-line R-CHOP chemotherapy. These results are also consistent with results from another study, evaluating the outcome of patients treated with R-VACOPB (rituximab/etoposide/ leucovorin/doxorubicin/cyclophosphamide/vincristine/prednisone/bleomycin) or R-CHOP, where the 5 year progression-free survival (PFS) was 81%. Further, in contrast to DLBCL, late relapses are uncommon, with a plateau in the survival curve beyond 2 years. Rarely, central nervous system relapses, typically parenchymal, can occur and it is unknown if the routine use of rituximab has reduced this risk.

The International Prognostic Index (IPI) which incorporates age (greater than 60), elevated LDH, advanced stage (III or IV), performance status (greater than two) and number of extranodal sites, was originally developed in diffuse large cell lymphoma, prior to the recognition of PMBCL. Thus, although there may have been cases of PMBCL included in this analysis, they were not evaluated as a separate group. Subsequent studies evaluating the IPI or the aaIPI (age-adjusted) have been discrepant in PMBCL. This may in part, reflect differences between studies assigning patient as stage IV or stage 2E if multiple but contiguous extranodal sites are involved.

In the few studies that have found the index useful, it has been the IPI that was applied, suggesting that it is primarily age that drives the poor prognosis. Even if the age-adjusted IPI is used, which eliminates the number of extranodal sites as a risk factor, most patients will have an elevated LDH, again reducing its discriminatory power. Other factors in individual studies that have been found to have a prognostic relevance include pleural and/or pericardial effusion, and poor performance status.

"What if" scenarios.

Patient presents with superior vena cava syndrome

If a patient presents with signs and symptoms of SVC syndrome, treatment needs to be urgently organized.

Ideally, a mediastinal biopsy should be organized prior to the institution of any therapy, however, if it is not felt to be safe to do so, steroids can be started, but the biopsy should be quickly organized as the diagnostic yield will be lowered, due to the high sensitivity of lymphoma to steroids. In addition, in patients awaiting biopsy confirmation who do not respond adequately to steroids, several doses of radiotherapy may be used prior to initiating definitive chemotherapy.


The presumed cell of origin of PMBCL is a thymic medullary, asteroid B-cell; however, the mechanism by which disease occurs is unknown. Molecular studies have shown distinctive chromosomal gains, including at 9p, which in some cases correlates with amplification of Janus kinase 2 (JAK2). Interestingly, mutations or deletions have been found in 45% of the SOCS-1 gene (suppressor of cytokine signaling), which is associated with increased JAK2 activity. SOCS1 mutations/deletions have also been found in 40% of CHL, supporting that this oncogenic pathway may be important in both diseases.

Recently, CIITA (master regulator of major histocompatability complex) breaks were found to occur in 38% of patients with PMBCL, and found to be associated with outcome in CHOP (like) treated patients. Down-regulation of surface HLA class II and overexpression of PD-1 (programmed cell death) were observed in CIITA gene fusions, raising the possibility that the rearrangement results in tumor cell immune escape. Similarly, CIITA breaks were also found in 15% of CHL.

Overall, the above observations support that PMBCL may be pathogenetically related to classical Hodgkin Lymphoma (cHL). Evidence of an overlapping relationship is further supported by gene expression profiling studies, that demonstrated that the molecular signature of PMBCL is not only distinct from that of DLBCL, but also has a striking resemblance to the expression profile of HRS cell lines. The signature is characterized by high expression of genes involved in cytokine and interleukin 4/13. Co-stimulatory molecules of the B7 family (CD80, CD86 and PDL-2 (programmed cell death ligand 1) were highly expressed in PMBCL and are evident in HRS cells.

Both PMBCL and cHL exhibit sclerosis, highlighting the importance of the extracellular matrix. Consistent with this, PMBCL demonstrated high levels of expression of the adhesion molecules and extracellular matrix components.

Further, as described, gray zone lymphomas (GZL) with overlapping features of CHL and PMBCL are recognized as distinct entity in the WHO. In addition, composite (CHL and PMBCL in the same tumor specimen) and sequential lymphomas (diagnosed with PMBCL and relapse with CHL, or vice versa) have been well described, and highlight the need for a repeat biopsy should the disease recur.

What other clinical manifestations may help me to diagnose primary mediastinal large B-cell lymphoma?

Important questions to ask on history
  • Respiratory symptoms: cough, dyspnea, chest pain, facial swelling

  • Systemic symptoms: B symptoms (Fever, night sweats, weight loss (greater than 10% of body weight), fatigue

  • Pruritis (rare)

  • Lumps: supraclavicular lymphadenopathy, chest wall mass

Important or unusual signs on physical examination
  • Distended veins over the neck and chest wall

  • Supraclavicular/cervical lymphadenopathy

  • Chest wall mass

  • Pemberton’s maneuver: occurs with superior vena cava compression from a mediastinal mass

– Development of facial flushing, distended neck and head superficial veins, inspiratory stridor and elevation of the JVP (jugular venous pressure) upon raising the patient’s arms above his/head.

What other additional laboratory studies may be ordered?


What’s the evidence?

Zinzani, PL, Martelli, M, Bertini, M, Gianni, AM, Devizzi, L, Federico, M. “Induction chemotherapy strategies for primary mediastinal large B-cell lymphoma with sclerosis: a retrospective multinational study on 426 previously untreated patients”. Haematologica. vol. 87. 2002. pp. 1258-64. [Retrospective analysis of dose-intensive regimens versus CHOP in PMBCL.]

Rieger, M, Osterborg, A, Pettengell, R, White, D, Gill, D, Walewski, J. “Primary mediastinal B-cell lymphoma treated with CHOP-like chemotherapy with or without rituximab: results of the Mabthera International Trial Group study”. Ann Oncol.. vol. 22. 2011 Mar. pp. 664-70. [Retrospective analysis of PMBCL patients enrolled on German aggressive lymphoma studies intended to evaluate the impact of rituximab.]

Savage, KJ, Al-Rajhi, N, Voss, N, Paltiel, C, Klasa, R, Gascoyne, RD. “Favorable outcome of primary mediastinal large B-cell lymphoma in a single institution: the British Columbia experience”. Ann Oncol.. vol. 17. 2006. pp. 123-30. [Retrospective study of PMBCL patients and comparison of outcome using dose-intensive regimens in comparison to CHOP.]

Dunleavy, K, Pittaluga, S, Janik, J, Grant, N, Steinberg, S, Staudt, L. “Primary mediastinal large B-cell lymphoma (PMBL) outcome is significantly improved by the addition of rituximab to dose adjusted (DA)-EPOCH and overcomes the need for radiation”. Blood.. vol. 106. 2005. pp. 929a[Phase II study of DA-EPOCH-R in PMBCL.]

Vassilakopoulos, TP, Pangalis, GA, Katsigiannis, A, Papageorgeiou, SG, Constantinou, N, Terpos, E. “Rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone with or without radiotherapy in primary mediastinal large B-cell lymphoma: The emerging standard of care”. Oncologist. vol. 17. 2012. pp. 239-49. [Large retrospective study of R-CHOP in PMBCL.]

Savage, KJ, Monti, S, Kutok, JL, Cattoretti, G, Neuberg, D, De Leval, L. “The molecular signature of mediastinal large B-cell lymphoma differs from that of other diffuse large B-cell lymphomas and shares features with classical Hodgkin lymphoma”. Blood. vol. 102. pp. 3871-9. [Study comparing the gene expression profiling signature of PMBCL, versus DLBCL with development of a gene expression based molecular diagnosis of PMBCL. A separate comparison to cell lines derived from CHL highlights the similarity between PMBCL and CHL.]

Rosenwald, A, Wright, G, Leroy, K, Yu, X, Gaulard, P, Gascoyne, RD. “Molecular diagnosis of primary mediastinal B cell lymphoma identifies a clinically favorable subgroup of diffuse large B cell lymphoma related to Hodgkin lymphoma”. J Exp Med. vol. 198. 2003. pp. 851-62. [Study comparing the gene signature of PMBCL to DLBCL, as well as to primary samples of classical Hodgkin lymphoma. These studies are complementary and suggest an overlapping pathogenetic overlap between PMBCL and CHL.]